Keras：使用 model.train_on_batch() 和 model.fit() 获得不同的精度。可能是什么原因以及如何解决这个问题？

Question

我有两个具有相同参数的相同模型。这两个都是在 MNIST 数据集上训练的。第一个使用 model.fit() 训练，第二个使用 model.train_on_batch() 训练。第二种模型的准确性较低。我想知道可能是什么原因以及如何解决它？

数据准备：

batch_size = 150
num_classes = 10
epochs = 12

# input image dimensions
img_rows, img_cols = 28, 28

# the data, split between train and test sets
(x_train, y_train), (x_test, y_test) = mnist.load_data()

if K.image_data_format() == 'channels_first':
    x_train = x_train.reshape(x_train.shape[0], 1, img_rows, img_cols)
    x_test = x_test.reshape(x_test.shape[0], 1, img_rows, img_cols)
    input_shape = (1, img_rows, img_cols)
else:
    x_train = x_train.reshape(x_train.shape[0], img_rows, img_cols, 1)
    x_test = x_test.reshape(x_test.shape[0], img_rows, img_cols, 1)
    input_shape = (img_rows, img_cols, 1)

x_train = x_train.astype('float32')
x_test = x_test.astype('float32')
x_train /= 255
x_test /= 255
print('x_train shape:', x_train.shape)
print(x_train.shape[0], 'train samples')
print(x_test.shape[0], 'test samples')

# convert class vectors to binary class matrices
y_train = keras.utils.to_categorical(y_train, num_classes)
y_test = keras.utils.to_categorical(y_test, num_classes)

模型 1：

model = Sequential()
model.add(Conv2D(32, kernel_size=(3, 3),
                 activation='relu',
                 input_shape=input_shape))
model.add(Conv2D(64, (3, 3), activation='relu'))
model.add(Conv2D(128, (3, 3), activation='relu'))
model.add(Conv2D(256, (3, 3), activation='relu'))
model.add(Conv2D(128, (3, 3), activation='relu'))
model.add(Conv2D(64, (3, 3), activation='relu'))
model.add(Conv2D(64, (3, 3), activation='relu'))
model.add(Conv2D(32, (3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
model.add(Flatten())
model.add(Dense(128, activation='relu'))
model.add(Dropout(0.5))
model.add(Dense(num_classes, activation='softmax'))

model.compile(loss=keras.losses.categorical_crossentropy,
              optimizer=keras.optimizers.Adadelta(),
              metrics=['accuracy'])

model.fit(x_train, y_train,
          batch_size=batch_size,
          epochs=epochs,
          verbose=1,
          validation_data=(x_test, y_test))
score = model.evaluate(x_test, y_test, verbose=0)
print('Test loss:', score[0])
print('Test accuracy:', score[1])

模型 1 精度：

测试损失：0.023489486496470636 测试精度：0.9924

模型 2：

model2 = Sequential()
model2.add(Conv2D(32, kernel_size=(3, 3),
                 activation='relu',
                 input_shape=input_shape))
model2.add(Conv2D(64, (3, 3), activation='relu'))
model2.add(Conv2D(128, (3, 3), activation='relu'))
model2.add(Conv2D(256, (3, 3), activation='relu'))
model2.add(Conv2D(128, (3, 3), activation='relu'))
model2.add(Conv2D(64, (3, 3), activation='relu'))
model2.add(Conv2D(64, (3, 3), activation='relu'))
model2.add(Conv2D(32, (3, 3), activation='relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Dropout(0.25))
model2.add(Flatten())
model2.add(Dense(128, activation='relu'))
model2.add(Dropout(0.5))
model2.add(Dense(num_classes, activation='softmax'))

model2.compile(loss=keras.losses.categorical_crossentropy,
              optimizer=keras.optimizers.Adadelta(),
              metrics=['accuracy'])

batch_size2 = 150
epochs2 = 12
step_epoch = x_train.shape[0] // batch_size2

def next_batch_train(i):
  return x_train[i:i+batch_size2,:,:,:], y_train[i:i+batch_size2,:]

iter_num = 0
epoch_num = 0
model_outputs = []
loss_history  = []

while epoch_num < epochs2:
  while iter_num < step_epoch:
    x,y = next_batch_train(iter_num)
    loss_history += model2.train_on_batch(x,y)

    iter_num += 1

  print("EPOCH {} FINISHED".format(epoch_num + 1))
  epoch_num += 1
  iter_num = 0 # reset counter


score = model2.evaluate(x_test, y_test, verbose=0)
print('Test loss:', score[0])
print('Test accuracy:', score[1])

模型 2 精度：

测试损失：0.5577236003954947 测试精度：0.9387

Answer 1

差异的四个来源：

1. fit()shuffle=True默认情况下使用，这包括第一个纪元（以及随后的纪元）
2. 你不使用随机种子；在这里看到我的答案
3. 您有step_epoch多个批次，但要迭代step_epoch - 1；更改<为<=
4. 你的next_batch_train切片已经结束了；这是它正在做的事情与它需要做的事情：
   • x_train[0:128] --> x_train[1:129] --> x_train[2:130] --> ...
   • x_train[0:128] --> x_train[128:256] --> x_train[256:384] --> ...

为了补救，您应该在您model2的火车循环中包含一个改组步骤- 或使用fitwith shuffle=False（不推荐）。另外，提示：64, 128, 256, 128, 64Conv2D 过滤器是一个非常糟糕的安排；你正在做的是大幅上采样，从某种意义上说“制造数据” - 如果你要使用更多的过滤器，也要strides按比例增加它们，以便层之间的总张量大小保持不变（或更少）。

下面所有提到的修复 + 更新的种子功能；运行 1 个 epoch，12 需要太长时间 - 如果 1 有效，那么 12 也可以。如果您愿意，可以保留您的原始模型，但我建议使用下面的一个进行测试，因为它要快得多。

import tensorflow as tf
import numpy as np
import random

def reset_seeds():
    np.random.seed(1)
    random.seed(2)
    if tf.__version__[0] == '2':
        tf.random.set_seed(3)
    else:
        tf.set_random_seed(3)
    print("RANDOM SEEDS RESET")

reset_seeds()
model = Sequential()
model.add(Conv2D(32, kernel_size=(3, 3),
                 activation='relu',
                 input_shape=input_shape))
model.add(Conv2D(64, (3, 3), activation='relu'))
model.add(Conv2D(32, (3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
model.add(Flatten())
model.add(Dense(128, activation='relu'))
model.add(Dropout(0.5))
model.add(Dense(num_classes, activation='softmax'))

model.compile(loss=keras.losses.categorical_crossentropy,
              optimizer=keras.optimizers.Adadelta(),
              metrics=['accuracy'])

def next_batch_train(i):
  return (x_train[i*batch_size2:(i+1)*batch_size2,:,:,:], 
          y_train[i*batch_size2:(i+1)*batch_size2,:])

iter_num = 0
epoch_num = 0
model_outputs = []
loss_history  = []

while epoch_num < epochs2:
  while iter_num < step_epoch:
    x,y = next_batch_train(iter_num)
    loss_history += model2.train_on_batch(x,y)

    iter_num += 1

  print("EPOCH {} FINISHED".format(epoch_num + 1))
  epoch_num += 1
  iter_num = 0 # reset counter

score = model2.evaluate(x_test, y_test, verbose=0)
print('Test loss:', score[0])
print('Test accuracy:', score[1])

更好的选择：使用改组

class TrainBatches():
    def __init__(self, x_train, y_train, batch_size):
        self.x_train=x_train
        self.y_train=y_train
        self.batch_size=batch_size

        self.indices = [i for i in range(len(x_train))]
        self.counter = 0

    def get_next(self):
        start = self.indices[self.counter] * self.batch_size
        end   = start + self.batch_size
        self.counter += 1
        return self.x_train[start:end], self.y_train[start:end]

    def shuffle(self):
        np.random.shuffle(self.indices)
        print("BATCHES SHUFFLED")

train_batches = TrainBatches(x_train, y_train, batch_size)

while epoch_num < epochs2:
  while iter_num <= step_epoch:
    x, y = train_batches.get_next()
    loss_history += model2.train_on_batch(x,y)

    iter_num += 1

  train_batches.shuffle()
  train_batches.counter = 0
  print("EPOCH {} FINISHED".format(epoch_num + 1))
  epoch_num += 1
  iter_num = 0 # reset counter

请注意，这并不能保证您的结果会与 一致fit()，因为fit()可能会以不同的方式洗牌（即使使用随机种子） - 但实际上实现是正确的。以上也不会在第一个时代洗牌（易于更改）。